According to a report by the Financial Times, OpenAI has disbanded its preparedness team as of the end of last month. This specialized unit was tasked with the critical mission of assessing whether new AI models posed existential or serious risks—such as the ability to facilitate cyberattacks or biological threats—and developing mitigation strategies.

Redistribution of Safety Duties

Instead of eliminating safety protocols entirely, the company has chosen to redistribute these responsibilities. According to the FT, duties related to specific high-stakes areas, including biological (bio) and cybersecurity risks, have been split and integrated into existing operational teams. This shift marks a move away from a centralized safety oversight model toward a more decentralized approach.

Why This Matters

BozokMedia analysis shows that this restructuring is part of a broader trend within OpenAI to transition from a research-heavy organization to a commercially driven powerhouse. As the company heads toward a highly anticipated IPO, the tension between rapid product deployment and rigorous safety testing is reaching a breaking point.

The fragmentation of specialized safety oversight could lead to blind spots in catastrophic risk assessment.

Historically, OpenAI has positioned itself as a leader in AI alignment and safety. However, the recent dissolution of key research-led departments suggests a pivot toward scaling and commercial viability, mirroring the intense pressure of the global AI arms race.

Frequently Asked Questions

1. What was the role of the preparedness team?
The team was responsible for identifying and mitigating severe risks posed by AI models, including rogue behavior and hacking capabilities.

2. Is OpenAI becoming less safe?
While OpenAI claims safety is being integrated into all teams, critics argue that losing a dedicated oversight body reduces the depth of risk analysis.